1. Field of the Invention
The present invention relates to a method for attending occurrence of a failure in a fixed-length cell exchange system, which exchanges cells having a fixed-length, or the like, an interface unit and a concentrator both equipped in such a fixed-length cell exchange, and the fixed-length cell exchange system.
This invention also relates to a matching system for use when software on a call control processor does not match with a status of a hardware including a firmware of a communication path system due to restarting or the like of the call control processor.
2. Description of the Related Art
ATM (Asynchronous Transfer Mode) exchange technology has been recommended and accepted as the next-generation exchange system, and has been aggressively studied by various organizations as the promising technology to accomplish a B-ISDN (Broadband-Integrated Services Digital Network).
FIG. 1 is a block diagram showing the schematic structure of an ordinary ATM exchange system. As shown in FIG. 1, an ordinary ATM exchange system comprises an ATM exchange 100, interface units 201 to 203 like UNI (User Network Interface) or NNI (Network Node Interface), and concentrators (or distribution units) 301 to 303 which concentrate (or distribute) the inputs and outputs of the respective interface units 201-203. Those interface units 201-203 are connected to the ATM exchange 100 via the concentrators 301-303, respectively.
The ATM exchange 100 is a switching section which deals with fixed-length data called as ATM cell having a fixed-length (normally, 53-byte length data consisting of a 5-byte header, which is a control information portion storing a transfer destination of the data or the like and a 48-byte data portion), and the switching section transmits and receives ATM cells via the interface units 201-203, and is designed to be able to a synchronously transfer (exchange) each ATM cell to the associated transfer destination in accordance with the header affixed to that ATM cell.
As shown in FIG. 1, therefore, the ATM exchange 100 normally comprises an ATM switch (exchange switch) 101, the so-called self-routing switch, which automatically determines the transfer path of an ATM cell, a main controller (CC) 102, which performs various controls of the components of the ATM exchange 100 including this ATM switch 101, and a main memory 103 which retains software for this main controller 102.
In this ATM exchange system, an ATM cell which is sent from a terminal accommodated in the interface unit 201 is input via the concentrator 301 to the ATM exchange 100 where the ATM switch 101 automatically selects the cell transfer path in accordance with the header affixed to this ATM cell and the ATM cell is then sent to the terminal or the like of the transfer destination via the concentrator 302, the interface unit 202 (or the concentrator 303 and the interface unit 203).
When a failure occurs in, for example, the ATM layer (cell multiplex/exchange layer) to the interface unit 201 in this ATM exchange system, the failure is detected by the interface unit 201, which in turn sends out an AIS (Alarm Indication Signal) cell of VP (Virtual Path)/VC (Virtual Channel) in the downstream direction of the ATM cell (toward the ATM switch 101), as indicated by the broken lines in FIG. 1, thereby informing the downstream stations on the side of the interface units 202 and 203 of the occurrence of the failure.
When the failure is cleared, the associated interface unit 201 stops sending the AIS cell to inform the downstream stations of the clearing of the failure.
While the above ATM exchange system can inform the downstream stations of the occurrence of a failure in an ATM layer (cell multiplex/exchange layer), there is no established way, at present, of informing the downstream stations of a failure if occurred in a line layer or the exchange switch section (ATM switch 101). Therefore, the latter failure cannot be detected by the downstream stations.
Even in the case where a failure in the ATM layer can be informed to the downstream stations, if some kind of an abnormality like a device failure occurs in the unit which detects the ATM layer failure and sends out the AIS cell (e.g., the interface unit 201), the AIS cell cannot be sent out, disabling the proper informing of the downstream stations of the ATM layer failure and the clearing of such a failure.
The ATM exchange system is generally provided with charge counters (cell counters) on the respective interface units 201-203 to charge subscribers. When a failure occurs in a line layer, an ATM layer, any of the interface units 201-203, the ATM switch 101 or the like, the number of input/output cells to the charge counter on each of the interface units 201-203 differs from the proper one, so that the accurate cell charge cannot be accomplished.
The B-ISDN based ATM exchange system, unlike the N-ISDN (Narrowband-Integrated Service Digital network) exchange system, can set a vast amount of logic paths to a single interface. For instance, the N-ISDN can set merely two channels (B/D) on the 64 Kbps interface, whereas the ATM exchange system can set a total of 2.sup.24 paths on a single interface by the virtual path identifier (VPI)=8 bits and virtual channel identifier (VCI)=16 bits according to the user-network interface (UNI) or the communication protocol which is defined by the demarcation point between a user's equipment (terminal or the like) and the network equipment.
With regard to a network-node interface (NNI) which is the interface for network nodes such as an exchange apparatus and a multiplexing apparatus, a total of 2.sup.28 paths can be set by VPI=12 bits and VCI=16 bits.
From the global viewpoint of the system, the number of paths should further be multiplied by the number of interfaces the system can accommodate.
When a call control processor which was disabled due to some kind of a failure or the like has been restored and has resumed its operation, the software status and the hardware (firmware) may not match with each other. In this sense, some kind of matching process should be performed.
When there are an incomparable number of connections as compared with that for the N-ISDN (which involves about two connections per interface), the matching between the software and hardware statuses after the restarting of the system takes time, which may result in a considerable time of service interruption.
Data that should be matched between the software and hardware include connection information (the input VPI/VCI, output VPI/VCI, tag information, usage parameter control (UPC) information, network parameter control (NPC) information, and the like) and failure information the hardware detects by itself (the virtual path alarm indication signal (VP-AIS), virtual channel alarm indication signal (VC-AIS) and the like). It is important to quickly match this information.
The usage parameter control or UPC is based on the overall capacity management which is performed even on a burst traffic in the ATM exchange in order to effectively use the network equipment. When a specific call causes a cell whose capacity exceeds the contracted capacity to flow, the communication quality of the entire network may be degraded.
A network which receives a call should therefore detect if the specified transfer capacity matches with the actual flow rate of cells, and executes a process of abandoning a cell which violates the rules when the actual traffic goes beyond the specified value for each contract with a subscriber. This control is called UPC.
The network parameter control or NPC is the control which is performed on a network node interface whereas UPC is performed on a user network interface.
The virtual path alarm indication signal or VP-AIS and the virtual channel alarm indication signal or VC-AIS are alarm signals that are sent to downstream nodes by a node which has detected an error to inform the detection of the error.
Each subscriber should contract a virtual channel (VC) between terminals with a network provider. This contract can allow each user network to be built up. This is called a virtual channel service (or ATM service).
In this case, the number of each virtual channel (VC) should be previously determined at the time of making a contract with the network provider, and the service of previously determining the number of a virtual channel (VC) is called a permanent virtual channel (PVC) service. The service in which connection to a virtual channel is established each occasion is called a switched virtual channel (SVC) service.
When a serious failure occurs in this service, the use of the temporary memory cannot be guaranteed so that everything should be released for an SVC call. Since over charging should be avoided for an SVC call, releasing should be carried out quickly.
With regard to a PVC which is assigned to an emergency call or the like with priority, unless nothing is wrong with the switch system, the service should continue as long as possible. For a light failure, however, the environment of the execution of a task cannot be inherited while the use of the temporary memory is guaranteed.
Under this situation, if the call control processor restarts the processing after the detection of a change in the hardware status before the software status is changed, failure information detected by the hardware itself cannot be recognized by the software. In this case, the software performs the normal process.
If the software is restarted while the hardware is detecting the restoration of a failed unit, the software cannot recognize the restoration and is thus unable to receive a new call endlessly.
In addition, as the B-ISDN unlike the N-ISDN has innumerable logic paths on each interface line, an inquiry about the state of each line by the software takes a considerable time until the service can become available after the call control processor has restarted the process. In this respect, the technique of shortening the time needed for the matching process is essential.
At present, however, even in an ATM exchange, actually, software sends data it holds to hardware (firmware) based on the N-ISDN for matching of pass information or matching of failure information of a virtual path VP or a virtual channel VC for each interface, and no consideration is given to the system which can shorten the matching time.